Struct std::random::WeightedIndex

Weighted discrete distribution.

Returns an index from 0..weights.len() with probability proportional to the weight at that index.

Sampling from this distributions has a complexity of O(log n) where n is the number of weights.

This distribution allocates a vector of cumulative weights on construction, so it needs to be freed when it is no longer needed.

Example

use std::random::{thread_rng, WeightedIndex};

let values = ["Mayo", "Mustard", "Ketchup", "BBQ Sauce"];
let dist = WeightedIndex::new(&[1, 0, 1, 2]);
defer dist.free();

// Prints "BBQ Sauce" with probability 1/2, and "Ketchup" or "Mayo"
// with probability 1/4 each
println!("{}", values[dist.sample(thread_rng())]);
Run this example

Methods

• fn new(slice: &[T]) -> WeightedIndex<T>

  source

  Creates a new weighted index distribution from a slice of weights.

  Panics if the slice is empty, if any of the weights are negative, or if the total weight is zero.

• fn from_iter<I>(iter: &mut I) -> WeightedIndex<T>

  I: Iterator<I, T>

  source

  Creates a new weighted index distribution from an iterator of weights.

  Panics if the iterator is empty, if any of the weights are negative, or if the total weight is zero.

• fn sample<R>(self: &WeightedIndex<T>, rng: &mut R) -> usize

  R: Rng<R>

  source

  Samples a distribution using a provided random number generator.

• fn move(self: &mut WeightedIndex<T>) -> WeightedIndex<T>

  source

  Creates a copy of the object, leaving an inert object behind.

  See Movable for details.

• fn free(self: &mut WeightedIndex<T>)

  source

  Frees the memory backing the object.

Mixins

• mixin<R> Distribution<WeightedIndex<T>, usize, R>

  R: Rng<R>

• fn sample_iter(self: &Self, rng: &mut R) -> DistIterator<Self, T, R>

  source

  Raturns an iterator producing an infinite stream of values sampled from the distribution.

  Example

  use std::random::{thread_rng, UniformInteger};
  
  let dist = UniformInteger::new(0..=5);
  
  // Prints 10 random numbers between 0 and 5 inclusive
  for i in dist.sample_iter(thread_rng()).take(10) {
      println!("{}", i);
  }
  Run this example